This application claims the priority of 101 48 325.2, filed 29 Sep. 2001, the disclosure of which is expressly incorporated by reference herein.
The invention relates to a central node for a data bus system having a bus monitor unit with a receiver for registering the signals on the data bus, and evaluation means which detect a faulty communication on the data bus, and at least temporarily block communication by the user causing the faulty communication, or compensate an incorrect communication that is triggered by interference influences. The bus monitor unit has timing means which are triggered in response to time patterns of the data bus for the transmission of each user, and detect a faulty communication when a user transmits outside the time patterns provided for it. The bus monitor unit is integrated into the central bus node and a plurality of bus branches of different users are combined at the central node so that the bus monitor unit can check a plurality of users for incorrect communication.
In contemporary means of transportation such as motor vehicles, trucks and aircraft, data bus systems are used for transmitting data between individual users, such as controllers, sensors and actuators. For this purpose, asynchronous data bus systems (for example the CAN bus) are used, in which messages are transmitted by the users as they occur. In addition, synchronous data buses are used in which each user is allocated one or more time slots in which it can transmit. The advantage of the latter data buses is that message collisions cannot occur when the system is functioning correctly. Time slots for transmission are assigned to each user by means of what are referred to as time patterns, and all the users on the data bus synchronize themselves with such a pattern, being able to transmit only in the time slot provided for them. The reception of messages is independent of this and the messages provided for each user are received at any time. Modern data buses, for example the MOST or the FlexRay data bus, have both synchronous and asynchronous phases.
In such data buses, communications faults may occur for different reasons. For example, a subscriber transmits on the data bus outside the time slot provided for it, in which case messages to be received are disrupted and a collision with other transmitted messages occurs. In order to block such incorrect communications on the data bus, or to compensate for the interference influences, bus monitor units are provided which are triggered in response to the time patterns of the data bus for the transmission of a user in order to detect an incorrect communication of a user when it transmits outside the time patterns provided for it.
German patent document DE 37 26 742 A1 discloses a bus monitor unit which intervenes in a user line between a user and data bus and interrupts it by means of two switches when a malfunction of the user occurs. Such a decentralized bus monitor unit is assigned to each of the individual users, so that each requires its own bus monitor unit which switches it off when a malfunction is detected. In data bus systems with up to 60 control units per vehicle and other sensors which are connected directly to the data bus, this gives rise to a situation in which a considerable number of bus monitor units, each assigned to the user, are provided within the means of transportation.
International patent document WO 01/13230 A1 discloses a method for blocking a user when incorrect communication occurs. For this purpose, “distributor units” are provided which, depending on the known customary transmission behavior of the users, permit transmission only within the statically assigned time slots. These distributor units, which are provided separately from the users and have a central bus monitor unit, observe one another mutually with respect to their functions and also carry out clock synchronization in order to synchronize with the time patterns of the data bus system. Each of the two disclosed distributor units has an integrated bus monitor unit. Each bus monitor unit forcibly causes each user to transmit to the other users only within its statically assigned time slots, based on its known transmission behavior. For this purpose, a time-registering means is provided which registers the time patterns of the data bus for the transmission of a user and, triggered by these time patterns, assigns a transmission slot to each user. Each distributor unit is connected via further communication channels, which are not associated with the data bus, to an external maintenance computer which can determine the parameters of the distributor unit and continuously monitors the correct functioning of the distributor unit. For comprehensive diagnostics, the external maintenance computer requires additional time-registering means which are again synchronized with the time patterns. As a result, the maintenance computer is either complex and expensive or carries out only an impoverished level of diagnostics.
One object of the present invention is therefore to provide a bus monitor unit of the type described above, whose correct functioning can be monitored during its operation, and can be quickly switched off in the event of a malfunction. This should be possible especially if the data communication outside the central node has partially collapsed.
This and other objects and advantages are achieved by the apparatus according to the invention, which includes in addition to the bus monitor unit, a diagnostic unit that is integrated into the central node. The diagnostic unit monitors the regular retriggering of the bus monitor unit in response to the time patterns, so that the bus monitor unit can be deactivated in the absence of retriggering. Alternatively, or in addition, it is possible to provide a configuration means (in addition to the bus monitor unit in the central node) for configuring the communication for each user in accordance with a configuration list stored in the central node so that, in the event of a fault, further communication of each user is either released or blocked individually.
The bus monitor unit which is integrated into the central node is suitable for monitoring access of users to the data bus, without having to install the bus monitor unit in a decentralized controller for this purpose. The central bus monitor unit and the diagnostic unit can be used to prevent faulty access to the data bus by a user (for example of a controller). It is possible to use the central node as a stand-alone unit encapsulated within a housing with self-diagnostics in the means of transportation. The central node is equipped with an integrated diagnostic unit without connecting lines for that purpose, which run in the vehicle and can create EMC problems as the result of external electromagnetic sources. In particular, the electronics of a decentralized data bus user (for example a controller) can also be integrated into the central node, so that the necessary hardware is available for the diagnostic computer.
As the bus monitor unit is integrated into the central node, its clock synchronization unit (which is already present) can be used as a time-registering means. Also, by virtue of the integration of the bus monitor and the diagnostic unit, there is no need for a separate communications network between the central node and the diagnostics unit, such as is necessary in other systems. Because of the integrated diagnostic unit, onboard diagnostics takes place in the central bus node, which in turn makes the latter particularly failsafe.
Sequential data slots in which users (i.e., a central or decentralized node) transmits are provided on the data bus. At the start of each new time slot, a trigger signal is generated by the time-registering means, and the time patterns defined by the trigger signals are made available to the bus monitor unit and to the diagnostic unit. The bus monitor unit can therefore check the communications data transmitted by a user to determine whether the transmission data occurs in the correct time slot.
Based on the second time pattern made available to it, the diagnostic unit checks whether the bus monitor unit regularly retriggers in response to the time patterns by means of the trigger signals (that is, whether it is appropriately retriggered at the start of each time slot). In this manner, the diagnostic unit checks the operation the bus monitor unit arranged within the same central node, and blocks it when there is a malfunction (for example, in the absence of retriggering). This is necessary because a malfunction of the bus monitor unit in the central node (especially incorrectly timed triggering) can result in complete blocking of the transmission signals on the data bus.
For this reason, according to the invention, both the bus monitor unit and the bus monitor diagnostic unit are integrated centrally into the central node. As a result, the chronological monitoring of the bus monitor unit functions very well owing due to the spatial proximity. On the other hand, however, it is also possible to achieve savings in terms of separate time-registering means which are necessary when there is a diagnostic unit that is arranged separately from the node. The central node with the integrated diagnostic unit therefore forms a closed system which preferably also has fault-handling routines, so that the central node is operationally capable independently of external diagnostic units, and has its own fault detection means.
The diagnostic unit has a watchdog connected to the time-registering means, that receives the signals for regular retriggering of the bus monitor unit and is connected via a control line to the bus monitor in order to switch it off in the absence of retriggering. The bus monitor unit is triggered periodically (i.e., retriggered) to carry out its monitoring function synchronously with the cyclical time patterns. The trigger signals are generated on the data bus by means of a clock synchronization unit. Such triggering is checked by the time-registering means or the diagnostic unit, and is used in the bus monitor unit for switching off in the case of malfunctions of the users, and in the watchdog for monitoring the bus monitor unit. Alternatively, the trigger signal can also originate from different time-registering means.
The bus monitor unit can be activated by an external signal or a signal of a configuration means. Before the bus monitor unit is activated, each user can transmit. On the other hand, the configuration means can switch the bus monitor unit switched to an inactive state in particular time periods, so that all the users can transmit. In this phase, it is possible, for example, for asynchronous communication to take place on the data bus, with no time patterns allotted to the individual users. After the asynchronous phase, the bus monitor unit is again switched to the active state to monitor the synchronous time behavior during the transmission of the individual users.
The watchdog preferably makes direct use of the stored time patterns of the configuration means again. It is connected to the time-registering means, and receives the signals for regularly retriggering the bus monitor unit. The time patterns which are accessed by the watchdog and the bus monitor unit can be configured using the configuration means, in particular when the data bus is powered up. Here, retrigger times for the bus monitor unit are defined, and intervals for asynchronous data transmission without monitoring by the bus monitor unit are determined. The synchronization of the bus monitor unit and the trigger signal for the watchdog can be derived from an identical trigger signal. This trigger signal is made available by the time-registering means.
The watchdog monitors the cyclical synchronization of the bus monitor unit with the time patterns of the data bus, and switches the bus monitor unit to an inactive state when the trigger signal fails to occur, blocking or releasing the communication in a way which can be configured for all the users. At the same time, a fault message of the watchdog is output to a control unit (for example on a display of the vehicle), in order to signal a fault condition. If the synchronization or retriggering of the bus monitor unit was faulty for a specific time, the watchdog can block it for a specific time and initiate a fault-handling routine of the diagnostic unit. The diagnostic, information can also be transmitted by a diagnostic computer to a control center or workshop.
The configuration means, which is also integrated into the central node, predetermines the time patterns of the data bus, so that the bus monitor detects the chronological sequences during the transmission operations of the different users. The time patterns of the configuration means are taken into account in particular when the data bus system is powered up, and are made available to the watchdog for the monitoring of the bus monitor unit.
The central node and the bus monitor unit, together with the diagnostic unit, are provided in a node computer housing, so that the central node may be used as a stand-alone version. For this purpose, the central node is configured as an intelligent unit which independently detects, and handles, faulty synchronization of the bus monitor unit. Faulty behavior of the central node does not disrupt of other communications channels.
A faulty node or user is temporarily switched to a passive state (in terms of communication technology) and can participate actively in the communication again after, for example, a restart or a configurable time interval. The time patterns, for the data bus can be configured flexibly, using the integrated configuration means. Because the configuration means is connected to the diagnostic device within a housing, only one time-registering means is necessary for the central node with bus monitor unit, watchdog and diagnostic unit. This provides an intelligent central node which is compact and can be used as stand-alone unit. By virtue of the integration of the configuration means, the start-up behavior is independent of the communications protocol, and the start-up is speeded up and occurs with a performance comparable to that of the communications protocol start-up.
The bus monitor unit is connected via an interface to a communications computer of the central node, which loads and calculates the time patterns for the accepted transmission slots of the individual users. The interface is a component of the configuration means, which itself can in turn be configured as a user of the data bus, with receiver and transmitter units for connecting to the data bus. Consequently, as a user of the data bus, the configuration means can also be provided with information on the time pattern via the data bus itself. This also makes possible a download or flash procedure in which software is transmitted from an external control center onto the data bus of the vehicle in order to provide new time patterns. In this manner, it is also possible to download the necessary new time pattern onto the data bus system by means of a software download, even when manufacturing the vehicle or when equipping it with a further controller.
The watchdog is preferably a component of a time-registering unit within the central node. The watchdog checks the cyclical synchronization of the bus monitor unit with respect to the provided time patterns, and generates a fault message when the synchronization by means of the respective trigger signals does not occur. Moreover, when there is a fault message, the watchdog can block the bus monitor unit from intervening in the bus communication.
In terms of the topology, the data bus with the central node can have a star shape or cascaded star shape. In a data bus with a cascaded star shape, two star nodes are arranged in direct connection with one another, and a data bus line is provided for the exchange of signals between the star nodes. In contrast to International patent document WO 01/13230, in which a unidirectional line is provided between the star couplers, the data bus line between the star couplers is of a bidirectional design here as in the customary data bus. In this way, two star-shaped data bus systems can communicate with one another via the star couplers.
The central bus node can also be embodied as a gateway, which can connect two different bus systems to one another. In this case, the data bus signals of one data bus in the central node are converted into the messages of the other data bus system. There is also provision for various time patterns in each of the different data bus systems, so that the bus monitor unit can monitor two different time patterns, and the watchdog checks the correct functioning of the bus monitor unit with respect to the two different time patterns.
The central node according to the present invention has the advantage that a star coupler, the bus monitor unit, the diagnostic unit and, if appropriate, the configuration means for the time patterns are arranged within an independently acting central node and within a housing. A flexible configuration of the time patterns can be carried out by means of an input unit on the central node. Likewise, the diagnostics are integrated into the intelligent central node, enabling particularly effective fault detection to be carried out on the time patterns.
In one embodiment, for example, a bus driver can either transmit or receive at one time. (That is, only the transmitter or the receiver in a driver can ever be connected through to the data bus.) The activation of the transmission/reception switches can be carried out in the user itself on the basis of signal activity on the bus line. If a bus driver detects activity on the external bus line, it transmits this signal to the other users via the star point of the central node. Alternatively, the bus monitor unit can release the communications path for each user.
The blocking function of the switches must be periodically checked in order to prevent concealed faults. This can be done by monitoring the activity on the data bus. If there is activity at a time when the switch should be released, the switch is defective. This method permits rapid and detailed diagnostics to be performed on the star coupler.